2.2.28 · Physics › Fluid Mechanics
Real fluids bahut messy hote hain: viscosity, vortices, turbulence. Lekin agar hum viscosity ignore kar dein (inviscid) aur ye assume karein ki fluid elements spin nahi karte (irrotational), toh velocity field ek single scalar ka gradient ban jaata hai — jaise electric field kisi potential ka gradient hota hai. Isse poora flow Laplace's equation solve karne mein aa jaata hai, jo linear hai. Linear matlab hum simple flows ko saath add kar sakte hain (superposition) aur complex flows bana sakte hain. Yahi poora fayda hai: mushkil fluid problems Lego ban jaate hain.
Definition Inviscid & Irrotational
Inviscid : viscosity μ = 0 . Koi internal friction nahi, koi shear stress nahi.
Irrotational : vorticity ω = ∇ × v = 0 har jagah. Fluid elements translate aur deform karte hain lekin apne axis ke around rotate nahi karte.
KYUN irrotational matter karta hai: Vector calculus ka ek theorem kehta hai ki agar kisi field ka curl zero ho (simply-connected region mein), toh woh field kisi scalar ka gradient hota hai. Toh:
∇ × v = 0 ⟹ v = ∇ ϕ
jahan ϕ velocity potential hai. Teen velocity components ki jagah ek scalar aa jaata hai.
DO ideas ko combine karne ka TARIKA. Hamare paas hai:
Irrotational ⇒ v = ∇ ϕ
Incompressible ⇒ ∇ ⋅ v = 0
Pehle ko doosre mein substitute karo:
∇ ⋅ ( ∇ ϕ ) = 0 ⟹ ∇ 2 ϕ = 0
Ye step kyun? Gradient ka divergence hi Laplacian hota hai by definition. Toh potential flow = Laplace's equation solve karna, jo physics mein sabse zyada study ki gayi linear PDE hai.
Intuition Linearity = superposition kyun
Laplace's equation linear hai: agar ϕ 1 aur ϕ 2 dono ∇ 2 ϕ = 0 satisfy karte hain, toh ϕ 1 + ϕ 2 bhi karta hai. Kyunki v = ∇ ϕ , velocities bhi add hoti hain . Yahi engine hai jisse simple flows se complex flows build hote hain.
Definition Stream function
ψ
2D mein, ψ ko is tarah define karo:
u = ∂ y ∂ ψ , v = − ∂ x ∂ ψ
Constant ψ ki lines streamlines hoti hain (fluid unke saath flow karta hai).
KYUN define karte hain: continuity automatically satisfy ho jaati hai:
∇ ⋅ v = ∂ x ∂ u + ∂ y ∂ v = ∂ x ∂ y ∂ 2 ψ − ∂ y ∂ x ∂ 2 ψ = 0.
Aur irrotationality force karti hai ki ∇ 2 ψ = 0 bhi ho. Pair ( ϕ , ψ ) conjugate harmonic functions hain: ϕ = const aur ψ = const lines har jagah perpendicular hoti hain (Cauchy–Riemann relations).
∂ x ∂ ϕ = ∂ y ∂ ψ , ∂ y ∂ ϕ = − ∂ x ∂ ψ
Har ek ko sabse simple ϕ /ψ likh ke aur velocities read off karke derive karo.
Worked example Rankine half-body = Uniform flow + Source
KYUN: source fluid ko outward push karta hai; dur upstream uniform flow dominate karta hai; jahan dono balance hote hain, ek stagnation point banta hai, aur ek streamline solid-looking nose jaisi ban jaati hai.
ψ = U y + 2 π m θ = U r sin θ + 2 π m θ .
Ye step kyun? Hum bas dono ψ 's add kar rahe hain — legal hai kyunki Laplace linear hai.
Stagnation point: v = 0 set karo. x -axis par (θ = π ), u = U − 2 π x m = 0 ⇒ x s = − 2 π U m .
Kyun: source ka leftward push wahan rightward stream ko exactly cancel karta hai.
Worked example Flow over a cylinder = Uniform flow + Doublet
ψ = U r sin θ − 2 π κ r s i n θ = U sin θ ( r − 2 π U r κ ) .
Ye step kyun? Doublet strength choose karo taaki ψ = 0 ek circle par ho. Radius a set karo jahan a 2 = 2 π U κ ; tab ψ = 0 at r = a for all θ — woh circle ek streamline hai, yaani ek solid cylinder!
Surface speed: u θ r = a = − 2 U sin θ . Maximum speed 2 U upar/neeche (θ = ± 90° ); stagnation at θ = 0 , π .
Worked example Lifting cylinder = Uniform + Doublet + Vortex → Magnus lift
Ek vortex Γ add karo. Vortex ek side flow speed up karta hai, doosri side slow karta hai → pressure difference. Kutta–Joukowski theorem lift per unit span deta hai:
L ′ = ρ U Γ.
Kyun matter karta hai: yahi airfoil theory ki neenv hai — spinning balls curve karte hain, wings lift karte hain.
Common mistake Common errors ko steel-man karna
"Irrotational ka matlab fluid circles mein nahi jaata." Kyun sahi lagta hai: "rotation" circular motion jaisa lagta hai. Fix: free vortex perfect circles mein move karta hai phir bhi irrotational hai (∇ × v = 0 ) singular center ke alawa har jagah. Irrotational local spin of a fluid element (vorticity) ke baare mein hai, global path shape ke baare mein nahi .
"Superposition pressures bhi add karta hai." Kyun sahi lagta hai: velocities add hoti hain, toh pressure kyun nahi? Fix: Bernoulli v mein nonlinear hai (p + 2 1 ρ v 2 = const). Velocity fields add karo, phir pressure ek baar end mein total velocity se compute karo. Pressures kabhi add mat karo.
"Potential flow drag predict karta hai." Fix: body ke paas steady flow ke liye ye zero drag predict karta hai (d'Alembert's paradox) kyunki viscosity/wake nahi hai. Lift aur pressure distribution ke liye achha hai, drag ke liye useless hai.
Recall Feynman: 12-saal ke bachche ko samjhao
Socho ek syrup jisme koi stickiness nahi hai, itna smoothly flow karta hai ki koi bhi chhota piece kabhi top ki tarah spin nahi karta. Tab poore flow ko ek "height map" se describe kiya ja sakta hai — fluid hamesha is map par "downhill" roll karta hai. Kyunki map ka rule simple hai (bas add hota hai), tum ek "seedha nadi" map, ek "spreading fountain" map, aur ek "swirl" map le ke unhe ek doosre par stack kar sakte ho aur ek ball ya wing ke around flow invent kar sakte ho. Swirl ko ball-flow par stack karo aur tumne explain kar diya kyun ek spinning ball curve karta hai!
"PUSH-VD" building blocks ke liye: P otential exist karta hai kyunki flow irrotational hai; U niform, S ource/sink, V ortex, D oublet — aur ek sneaky H : doublet+uniform = cylinder tak H alf ka raasta. Yaad rakho "Velocities add hoti hain, Pressure nahi."
Potential flow define karne wale do assumptions kya hain? Inviscid (
μ = 0 ) aur irrotational (
∇ × v = 0 ).
Irrotationality humein velocity potential define karne kyun deti hai? Zero curl ⇒ field ek gradient hai:
v = ∇ ϕ .
Potential flow ki governing equation derive karo. Incompressible
∇ ⋅ v = 0 plus
v = ∇ ϕ gives
∇ 2 ϕ = 0 (Laplace).
Yahan superposition valid kyun hai? Laplace's equation linear hai, toh solutions ka sum bhi solution hai; velocities add hoti hain.
Stream function define karo aur uski key property batao. u = ∂ ψ / ∂ y , v = − ∂ ψ / ∂ x ; lines ψ = const streamlines hain; continuity automatically satisfy hoti hai.
Strength m ke source ke liye ϕ aur ψ kya hain? ϕ = 2 π m ln r , ψ = 2 π m θ , with u r = m / ( 2 π r ) .
Circulation Γ ke free vortex ke liye u θ kya hai? u θ = Γ/ ( 2 π r ) ; center ke alawa irrotational.
Radius a ke cylinder ke paas flow kaise build karte ho? Uniform flow + doublet with a 2 = κ / ( 2 π U ) ; circle r = a streamline ban jaati hai.
Non-spinning cylinder par surface speed kya hai? u θ = − 2 U sin θ ; upar/neeche max 2 U , aage/peeche stagnation.
Uniform+source (Rankine half-body) ka stagnation point kahan hai? x s = − m / ( 2 π U ) axis par, jahan source push stream ko cancel karta hai.
Kutta–Joukowski lift batao. L ′ = ρ U Γ per unit span; circulation se lift.
D'Alembert's paradox kya hai? Body ke paas steady potential flow zero drag deta hai (no viscosity, no wake).
Common error: kya pressures superpose hote hain? Nahi — velocity fields add karo, phir (nonlinear) Bernoulli ek baar apply karo.
Laplace's equation — woh linear PDE jo superposition ko kaam karne deti hai
Bernoulli's equation — superposed velocity field ko pressure mein convert karta hai
Vorticity and circulation — irrotationality aur vortex define karta hai
Kutta–Joukowski theorem — circulation se lift
Cauchy–Riemann equations — ϕ aur ψ ko conjugate harmonics ki tarah link karta hai
Magnus effect — spinning cylinder/ball lift
Conformal mapping — in flows ko airfoil shapes tak extend karta hai
Laplace equation nabla2 phi=0
Basic building-block flows